Caar

detect_columns() takes the same arguments as dict_from_file() and pickle_from_file(), plus the 'meta'=True argument in order to return column metadata in a dict, without records.

Row text delimiters may be specified by the user. (By default, commas, tabs, pipes and spaces are automatically detected, in that priority order). In addition, the user can indicate characters that surround the data fields and that should be removed. Double and single quotes are automatically detected and removed by default.

strftime= keyword argument can take string to specify unusual datetime formats, if dateutil is not able to infer it automatically.

Moved time column validation to sample records.

Each of the underlying modules works as described and documented.

The package has been tested with a test suite and a sample data set under Python 2.7, 3.4 and 3.5.

While new features may be added, the existing API is expected to undergo only minor (if any) changes.